Properties of Ackermann's function

In summary: Could I improve something? (Wondering)In summary, the conversation discusses the properties of Ackermann's function, including its value compared to its inputs, and how it relates to other functions. The conversation also includes discussions on how to formally prove these properties through induction on various variables.
  • #1
mathmari
Gold Member
MHB
5,049
7
Hey! :eek:

I want to show the following properties of Ackermann's function:

  1. $A(x,y)>y$.
  2. $A(x,y+1)>A(x,y)$.
  3. If $y_2>y_1$, then $A(x,y_2)>A(x,y_1)$.
  4. $A(x+1, y) \geq A(x,y+1)$.
  5. $A(x,y)>x$.
  6. If $x_2>x_1$, then $A(x_2, y)>A(x_1, y)$.
  7. $A(x+2, y)>A(x,2y)$.
I have done the following:

  1. Induction on $x$.
  2. Induction on $x$.
    Base case: For $x=0$ we have $A(0,y+1)=y+2>y+1=A(0,y)$, by the definition of $A$.

    Inductive hypothesis: We assume that it holds for $x=n$, that means $A(n,y+1)>A(n,y)$ (I.H.x)

    Inductive step: We want to show that $A(n+1, y+1)>A(n+1,y)$.
    By the definition of $A$ and by the property $1$ we have $A(n+1, y+1)=A(n,A(n+1, y))>A(n+1,y)$.

  3. Induction on $y$.
    Base case: For $y=0$ we have $A(x+1, 0)=A(x,1)=A(x,0+1)$.

    Inductive hypothesis: We assume that it holds for $y=m$, that means $A(x+1, m) \geq A(x,m+1)$ (I.H.)

    Inductive step: We want to show that $A(x+1, m+1) \geq A(x,m+2)$.
    By the definition we have $A(x+1, m+1)=A(x,A(x+1, m))$. By the (I.H.) and the property $1$ we have $A(x+1, m) \geq A(x,m+1)>m+1 \Rightarrow A(x+1, m) \geq A(x,m+1) \geq m+2$. By the property $3$ we have $A(x,A(x+1,m)) \geq A(x,m+2)$. So, we cocnlude that $A(x+1, m+1) \geq A(x,m+2)$.
  4. By the property $4$ we have $A(x+1, y) \geq A(x,y+1)$. So, we have $A(x,y) \geq A(x-1, y+1) \geq A(x-2, y+2) \geq \dots \geq A(0,x+y)=x+y+1>x$.
    But how can we prove this formally??

  5. Induction on $y$.
    Base case: For $y=0$ we have by the property $6$ $A(x+2, 0)>A(x,0)=A(x,2 \cdot 0)$.

    Inductive hypothesis: We assume that it holds for $y=m$, that means $A(x+2, m)>A(x,2m)$ (I.H.)

    Inductive step: We want to show that $A(x+2, m+1) > A(x,2(m+1))$.
    By the definition of $A$ we have $A(x+2, m+1)=A(x+1, A(x+2, m))$. By the (I.H.) we have $A(x+2, m)>A(x,2m)$. By the property $3$ we have $A(x+2, m+1)>A(x+1, A(x,2m))$. By the property $4$ we have $A(x+1, A(x,2m)) \geq A(x,A(x,2m)+1)$ According to the property $1$, $A(x,2m)>2m \Rightarrow A(x,2m) \geq 2m+1$ or equivalently $A(x,2m)+1 \geq 2m+2=2(m+1)$. By the property $3$ we have $A(x,A(x,2m)+1) \geq A(x,2(m+1))$. So, $A(x+2, m+1)>A(x,2(m+1))$.

Is this correct?? Could I improve something?? (Wondering)

Could you give me some hints to prove the properties $3, 6$ ?? (Wondering)
 
Physics news on Phys.org
  • #2
At the proof of the properties $3$ and $6$ do we use the properties $2$ and $4$ respectively?

But how? (Wondering)

For example, for the property $3$ would the proof be as follows?

Induction on $x$.

Base case: For $x=0$ we have $A(0,y_2)=y_2+1>y_1+1=A(0,y_1)$

Inductive hypothesis: We assume that it holds for $x=n$, that means $A(n,y_2)>A(n,y_1)$ (I.H)

Inductive step: We want to show that $A(n+1,y_2)>A(n+1,y_1)$.

How can we continue?
 
  • #3
I tried again to prove the property $3$:

$y_2>y_1 \Rightarrow y_2=y_1+d, d>0 \Rightarrow d=y_2-y_1>0$

Induction on $d$.

Base case: For $d=1$ we have $A(x, y_2)=A(x, y_1+1) \overset{(2)}{>}A(x, y_1)$.

Inductive hypothesis: We assume that it holds for $d=k$, that means that $A(x, y_2)=A(x, y_1+k)>A(x, y_1)$ (I.H.)

Inductive step: We want to show that $A(x, y_1+(k+1))>A(x, y_1)$
$A(x, y_1+k+1)=A(x, (y_1+k)+1)\overset{(2)}{>}A(x, y_1+k)\overset{ (I.H.) }{>}A(x, y_1)$ Is this correct? Could I improve something? (Wondering)

The proof of the property $6$ is similiar, isn't it?

On which variable do we use the induction for the property $5$ ?
 
  • #4
For the property $5$ do we do the following?

We will show that $$A(x, y) \geq A(0, x+y)$$ by induction on $k=x+y$. Base case: For $k=0$ we have $x=y=0$, so $A(0, 0) \geq A(0, 0)$.

Inductive hypothesis: We assume that it stands for $k=n$, $A(x, y) \geq A(0, n)$. (I.H.)

Inductive step: We will show that it stands for $k=n+1$, $x+1+y=n+1$, that means that $A(x+1, y) \geq A(0, n+1)=(n+1)+1=n+2$.
We have that $A(x+1, y) >A(x, y) \overset{ (I.H.) }{\geq} A(0, n)=n+1$.
$A(x+1, y)>n+1 \Rightarrow A(x+1, y) \geq n+2$. So, we have that $A(x, y) \geq A(0, x+y)=x+y+1>x+y$. Is this correct?
 

Related to Properties of Ackermann's function

What is Ackermann's function?

Ackermann's function is a mathematical function that is used to show the growth rate of certain algorithms. It was first introduced by Wilhelm Ackermann in 1928.

What are the properties of Ackermann's function?

The properties of Ackermann's function include being a total computable function, being a non-primitive recursive function, and having extremely fast-growing values as the inputs increase.

Why is Ackermann's function important?

Ackermann's function is important because it can be used to compare the efficiency of different algorithms and to analyze the growth rate of certain problems. It is also a fundamental example in the study of computability and complexity theory.

What is the time complexity of Ackermann's function?

The time complexity of Ackermann's function is extremely high, with a runtime that grows faster than any polynomial function. It is often used as a benchmark for comparing the efficiency of different algorithms.

Can Ackermann's function be computed for all inputs?

No, Ackermann's function is not defined for all inputs. It is only defined for non-negative integer inputs. As the inputs increase, the values of the function grow extremely quickly and eventually exceed the capacity of any computer to compute.

Similar threads

  • Set Theory, Logic, Probability, Statistics
Replies
1
Views
1K
Replies
0
Views
434
  • Set Theory, Logic, Probability, Statistics
Replies
2
Views
1K
  • Set Theory, Logic, Probability, Statistics
Replies
0
Views
1K
  • Set Theory, Logic, Probability, Statistics
Replies
1
Views
1K
  • Set Theory, Logic, Probability, Statistics
Replies
2
Views
1K
  • Set Theory, Logic, Probability, Statistics
Replies
2
Views
1K
  • Set Theory, Logic, Probability, Statistics
Replies
4
Views
889
  • Set Theory, Logic, Probability, Statistics
Replies
1
Views
984
  • Set Theory, Logic, Probability, Statistics
Replies
3
Views
2K
Back
Top